Pancreatic adenocarcinoma (PDA) carries one of the most dismal prognoses in all of medicine and is currently the 4th leading cause of cancer death in the United States. Several subtypes of PDA have been identified that vary greatly with regard to their lethality. Preliminary studies presented in this proposal support the notion that diverse PDA subtypes form from distinct pancreas cell populations, including duct and acinar cells. Importantly, we have identified a specific gene coding for a chromatin remodeling enzyme that modulates the response to tumor-inducing cues in these different cell types. Elimination of the gene in duct cells of transgenic mice supports the development of morphological and molecular changes associated with the formation of the more benign PDA subtype. In contrast, the gene product is required for the development of the more aggressive form of PDA in acinar cells. Thus, we have identified a novel regulator of PDA subtype formation. We propose to test how the loss of the chromatin remodeler changes the differentiation state of pancreatic duct cells and makes them prone to develop the more benign subset of pancreatic cancer. Conversely, we present an experimental strategy to test how the loss of this gene prevents formation of the more aggressive form of PDA from acinar cells. We have already identified potential target genes for both the aggressive and more benign subtypes and will test whether manipulation of these factors initiates or prevents the early stages of benign and aggressive tumor formation. Finally, we are focusing on one particular target that we recently characterized to be essential for the development of the aggressive form. In this application we will test whether changes in gene expression of that factor can interfere with the development of the PDA subtypes. Throughout the proposal, we are using state of the art transgenic animals and molecular tools to address the questions outlined above. Furthermore, we have established close collaborations with clinical colleagues both at UCSF and at the University of Utah who will provide us with specimen of human cancer samples to test and verify that new factors identified in our mouse models have relevance to the human disease. It is our goal to identify both novel diagnostic markers that can be used for the early detection of the subsets of PDA as well as novel therapeutic targets that could be exploited to attack the cancer ones it has formed. These studies fulfill a critical gap as our knowledge about the formation and molecular mechanisms underlying the development of the more benign form of pancreatic cancer is limited. By understanding how the more benign cancer forms and how this differs from the formation of the aggressive form, we hope to gain critical insights into how one can manipulate the aggressive tumors to reduce it growth potential.